The development of the methods of cancer control epidemiology has focused on these areas: 1) studies in causal and preventive inference, 2) studies in interactions, 3) the analysis of epidemiologic reasoning, 4) the ethics of prevention, and 5) design of cancer control research studies. Criteria for causal and preventive inference have been improved. Previous criteria fall into two categories: those dependent upon the specific form of the hypothesis and those independent of the form of the hypothesis. Two general criteria replace all existing criteria: predictability and testability. Studies of causal and preventive interactions have revealed that the links between biological and statistical models are necessary for scientific progress while the link between the discovery of interaction and its application requires examination of ethics. The use of the multiplicative model of preventive interaction (and the additive model of causal interaction) as thresholds for public health action are not justified on ethical grounds. Research underway includes: 1) Application of general criteria of epidemiologic inference to determine status of hypothesis that dietary alcohol causes breast cancer and dietary fiber prevents colon cancer, 2) Predicting magnitudes of effects from the sufficient component causes model, and examining the role of magnitude of association in making inferences, 3) Application of a new causal theory (epigenesis theory) to the evaluation of causal and preventive interactions, 4) Application of a new method of epidemiologic reasoning (mega-analysis) to: the use of case- control methods for screening, the use of negative results in causal inference, the debate regarding the use of p-values, confidence limits, and p-value functions, the role of complexity in causal inference, and the links between scientific, technologic and ethical reasoning. 5) A critical study of the ethics of epidemiology in advising for or against preventive interventions, and 6) Defining criteria and issues of phase IV (defined population) studies.